Somewhat belated but the last two BoMs for 2014 are atNovember, WhangareiandDecember,... [read more]
Seems that the demo uploaded was the wrong one. Sorry. 188.8.131.52 now... [read more]
Archie began as a very simple tool to help engineers use the concepts presented by Jack Heyman. Over the years it has advanced considerably. Initially it was developed as part of a research programme. It was not conceived as a tool for general users. Support and encouragement from the Science and Engineering Research Council (now EPSRC), and from potential users, led to the appointment of Fraser Smith as a research assistant to a Bill. Fraser took over most of the development work of Archie, introducing tools to aid the construction of more complex models and systems of error-checking etc that are so necessary in commercial software.
The first real user of Archie was the Northern Ireland Department of the Environment. Bill has a vivid memory of visiting them and discussing the use of Archie with one of their engineers. Will this give me answers? He asked. No, said Bill, it will help you to explore possibilities. At that point, Archie became a business.
The Archie programme was based directly on the mechanism approach used by Heyman. We very soon came under pressure to find a way of dealing with bridges of more than one span. Eventually, we realised that the way forward lay in more direct interaction between the engineer and the computer. What we produced was essentially a flexibility analysis. The multi programme was born.
The original Archie programme was built for MS-DOS. As computers became more powerful Windows became more common as an operating system and Archie-Win was developed around 1990. Multi never appeared as a Windows program.
In 1995 the original team was broken up when Bill moved to Exeter University. By that time, the program was entirely the responsibility of Fraser and was owned by the University of Dundee. Fraser continued to maintain the program but could no longer find the energy to develop it. In 1999, he handed it back to Bill who formed a relationship with Zoltàn Juhàsz and began a complete redevelopment of the program. Archie-M was completed in 2000 and the new company OBVIS limited was set up to market it.
Four hinges or three
Heyman's analysis was based on a structural mechanism with four hinges in the arch. In fact the arch first goes through a statically determined the state with only three hinges. In multi-span arches the 3 hinge model is much more powerful. With three hinges in each arch the structure is statically determinate. The user is then free to modify hinge positions interactively in order to explore the best possible configuration. The aim is not to show how the bridge truly responds to load, but rather how it might sustain the loads placed upon it. We describe this as an interactive flexibility analysis. We believe it is a very powerful concept offering opportunities for new paradigms in engineering software.
We have said that the understanding of arch bridges is at a remarkably low level. Progressively developing software which remains based on clearly erroneous models seems to us a bad investment of time. We are therefore involved, chiefly through Bill Harvey Associates Ltd, in continuous research and development of arch bridge behaviour. Some of that research is analytically based, using more complex software to explore the development of simpler models for use in future first level assessment. The research is often driven by new thoughts about bridge behaviour which developed when looking at particular bridges. This raises questions which can be explored by analysis or by measurement of elements of behaviour.
Recent analytical research has centred on the use of very simple finite element models to explore the interaction between an arch and the field and loads it supports and to study the flow of force in a large barrel. It is our firm belief that until models can successfully be directed at both skew and square arches in single and multiple spans we will not have an adequate tool for arch assessment.
A study of the MEXE method undertaken for the UIC highlighted the development of and errors in the distribution model which is fundamental to current practice in transforming a three-dimensional bridge into a two-dimensional model for analysis. This process clearly breaks down with skew bridges but it has only recently become obvious that it is quite inadequate and square bridges too. The effective strip model puts undue demand on abutments and piers. Our new model in which the live load thrust distributes through the bridge is going to require a new concept of bridge failure, but should allow us to analyse properly for such things as weak patches in an arch.
Bridge testing and monitoring
Testing and monitoring arch bridges often produces useless results at great expense. Over the past 10 years Bill has developed a range of equipment which can be deployed quickly and which can measure patterns of displacement under live traffic loads. The results from these tests, when combined with analytical work described above show that live load thrust truly does distribute. This helps to explain why viaducts and abutments behave so much better than crude analyses suggest.
Over 30 years in the business Bill has carried out many arch assessments. He has also reviewed very many more. It has become clear that the vast majority of assessments are based on very poor quality data. Inspecting and measuring arches for assessment is a skilled job. Indeed assessment itself, particularly analysis for assessment, is a matter of exploration rather than simply plugging data into a process. It is therefore vital that the basic data is correct but also that it is backed up with appropriate photographs drawings and notes to allow the analyst to understand the nature of the structure. Bill is happy to discuss these problems with users and also to run training courses to help reduce the occurrence of error.